Henry,
Thanks for looking into this - it was late last night and I wasn't as thorough
as Bill Lam was (thanks for the nudge Bill) and didn't look earlier - here is a
little update -
JVERSION
Engine: j805/j64/darwin
Release: commercial/2016-12-11T08:17:56
q=. (50e6 ?.@$ 10){'0123456789'
timex 'z=. rl q'
0.828379 603981056
Engine: j807/j64/darwin
Release-e: commercial/2019-11-04T12:57:58
q=. (50e6 ?.@$ 10){'0123456789'
timex 'z=. rl q'
1.973677 649991296
Engine: j901/j64avx2/darwin
Release-c: commercial/2020-01-11T12:24:26
q=. (50e6 ?.@$ 10){'0123456789'
timex 'z=. rl q'
6.426628 649995392
Above are run in a 2.7 GHz Intel Core i5
Below is another timing (on a very different machine).
But it at least gives your Release-c a factor of 2 faster (in real time ...
;-)
j504/2005-03-16/15:30
Running in: Linux
Linux version 2.4.20-28.8 (bhcomp...@daffy.perf.redhat.com) (gcc version 3.2
20020903 (Red Hat Linux 8.0 3.2-7)) #1 Thu Dec 18 12:53:39 EST 2003
q=. (50e6 ?.@$ 10){'0123456789'
timex 'z=. rl q'
12.9823 3.35545e8
$ cat /proc/cpuinfo
processor : 0 (the one and only core)
vendor_id : GenuineIntel
cpu family : 6
model : 5
model name : Pentium II (Deschutes)
stepping : 2
cpu MHz : 399.069
cache size : 512 KB
.....
So, the factor of 2 times faster is offset by the processor running at 6.75
times slower clock rate ;-)
I'm feeling a little remiss in not paying closer attention during the beta
process... 🤷♂️
> On 2020Jan 15, at 17:35, Henry Rich <henryhr...@gmail.com> wrote:
>
> Fixed for next release.
>
> Henry Rich
>
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
